In 10 alternating-polarity SEEG trains, the polarity-invariant response did not adapt more than the artifact component

Plan 464344c8 tested whether the polarity-invariant pulse-locked component S of 43.2 Hz alternating-polarity bipolar SEEG stimulation declined from the first to last second, while the polarity-flipping component D remained constant. In OpenNeuro ds006519 v1.0.0, 10 trains from 6 patients passed pulse checks; non-alternating 50 Hz trains were excluded. For each non-stimulated channel, RMS of S and D was measured 2–15 ms after pulses in the first and last train seconds. The primary patient-level S−D log-ratio was −0.053 (3/6 patients negative; exact sign-flip p = 0.50); S changed by a factor of 0.938 and D by 0.988. In the secondary analysis of channels with middle-second S greater than twice the noise floor, S changed by a factor of 0.860 and D by 1.007; S−D was −0.158 (4/5 patients negative; p = 0.125). An earlier secondary selection using first-second S gave p = 0.0625 but was biased by regression to the mean and was replaced. The artifact component was stable. Adaptation of S remains possible but was not shown, so this test does not determine the method assumption. The analysis includes only 6 patients and 10 trains; most patients contributed one train. Pulse alignment was to the nearest sample, amplifier saturation was not modelled, and the visual-effect dataset ds005169 was not analyzed. The package preserves the analysis code, results, source metadata, and chunk hashes.

stereo-EEG · intracranial stimulation · alternating polarity · pulse-locked response · artifact · neural adaptation · sign-flip test · OpenNeuro ds006519